jackaduma / cyclegan-vc2 Goto Github PK
View Code? Open in Web Editor NEWVoice Conversion by CycleGAN (语音克隆/语音转换): CycleGAN-VC2
License: MIT License
Voice Conversion by CycleGAN (语音克隆/语音转换): CycleGAN-VC2
License: MIT License
librosa.output was removed in librosa version 0.8.0. Soundfile could be used instead. Thanks.
Dear friend,
PS D:\CycleGAN-VC2-master\CycleGAN-VC2-master> python train.py
Traceback (most recent call last):
File "train.py", line 520, in
cycleGAN = CycleGANTraining(logf0s_normalization=logf0s_normalization,
File "train.py", line 106, in init
self.start_epoch = self.loadModel(restart_training_at)
File "train.py", line 442, in loadModel
checkPoint = torch.load(PATH)
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 595, in load
return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args)
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 774, in _legacy_load
result = unpickler.load()
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 730, in persistent_load
deserialized_objects[root_key] = restore_location(obj, location)
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 175, in default_restore_location
result = fn(storage, location)
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 151, in _cuda_deserialize
device = validate_cuda_device(location)
File "C:\Users\Administrator\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\serialization.py", line 135, in validate_cuda_device
raise RuntimeError('Attempting to deserialize object on a CUDA '
RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU.
but when i input the torch.cuda.is_available() under python cmd, the return value is true.
this issue block me for days,any suggestion will work?
cannot thank more.
训练之后是否可以将文本转为某个人音色的语音?
Hello, I just started to learn voice conversion.And I want to know how to write a demo by using this frame? How do I use another person’s voice to speak the content of the person’s speech with the voice in the training set?
Is it necessary to have a pre-trained model to start training? Can we train directly without a pretrained model?
I am checking the model provided in this repository. How to do the inference of the given model? I want to check the model using my dataset. How to generate the output by passing our wavfile?
FileNotFoundError: [Errno 2] No such file or directory: './model_checkpoint/_CycleGAN_CheckPoint'
请问换其他数据集测试的时候,报上述错误是怎么回事呀,求解答
I want to know the final loss of D and G
Is there any way to re-synthesis new audio using the trained model? I don't know whether I miss some info or the info such as synthesis.py
is left out.
Thanks in advance.
在自己的服务器上使用预训练模型跑了一周左右,但是得到的转换效果仍然不理想,得出的结果就像是A的语速和韵律变成了B的语速和韵律,但是总体听上去说话人的身份好像并没改变,只是语速和韵律和目标说话人的一样
请问这个模型跑多长时间才能得到比较好的效果,在训练过程中有没有什么技巧,能够让转换之后的声音效果更好
谢谢您
Hey, am I missing something or the second step adverserial loss is missing in this implementation?
请问这个_CycleGAN_CheckPoint是怎么才能有啊
如果我有A数据量很多,B数据量较少,通过B转换为A的音色,效果会怎样呢?
关于数据量的关系,您有相关的分析吗?例如ABCD数据量都较多,是否相互转换音质高一点呢?又例如A多B少,A转B和B转A效果各自会怎样?
男声和女声之间的转换效果如何?
怎样的声音转换的效果会好一点?
这个方案做多个说话人转为一个说话人有效果吗?
另外,Donation可以提供支付宝或微信二维码,**同学更熟悉。
https://arxiv.org/abs/2010.11672
Accepted to Interspeech 2020.
File "train.py", line 521, in
cycleGAN.train()
File "train.py", line 147, in train
for i, (real_A, real_B) in enumerate(train_loader):
File "C:\Users\mike\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.8_qbz5n2kfra8p0\LocalCache\local-packages\Python38\site-packages\torch\utils\data\dataloader.py", line 363, in next
data = self._next_data()
File "C:\Users\mike\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.8_qbz5n2kfra8p0\LocalCache\local-packages\Python38\site-packages\torch\utils\data\dataloader.py", line 403, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "C:\Users\mike\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.8_qbz5n2kfra8p0\LocalCache\local-packages\Python38\site-packages\torch\utils\data_utils\fetch.py", line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "C:\Users\mike\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.8_qbz5n2kfra8p0\LocalCache\local-packages\Python38\site-packages\torch\utils\data_utils\fetch.py", line 44, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "C:\Users\mike\Desktop\cycleganvc\trainingDataset.py", line 47, in getitem
assert frames_B_total >= n_frames
AssertionError
when trying to use my own training data
After I pre process my data and run the training I get this
however if I pre process the data you supplied and use that cache file it works
Any suggestions?
16khz mono 32 bit floats
你好!我最近在尝试复现ASSERT论文,但是在数据集预处理方面遇到了一些问题,我有看到你在issue中回答,可以咨询一下相关问题嘛?
非常期待您的解答,万分感谢!!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.